2013/1/17 Markos Chandras <hwoar...@gentoo.org>: > On 16 January 2013 20:09, Alexis Ballier <aball...@gentoo.org> wrote: >> On Wed, 16 Jan 2013 12:40:02 +0000 (UTC) >> "Tomas Chvatal (scarabeus)" <scarab...@gentoo.org> wrote: >> >>> scarabeus 13/01/16 12:40:02 >>> >>> Modified: ChangeLog >>> Added: ffmpeg-9.ebuild >>> Removed: ffmpeg-0.10.2-r1.ebuild >>> Log: >>> Add new virtual for 1.1/9 series. Masked. Also it has switched dep >>> order as will be announced upon unmasking. >> >> ... and since we are committing silently without any real discussion I >> will switch the dep order again and announce it much later without >> leaving room for discussion :) >> >> More seriously: Why ? Who decided this ? > > I agree. This is a big change so there should be a discussion about > this or at least an announcement that this is going to happen on the > Xth of February. Did you actually test that the tree is ready for > libav as the default ffmpeg provider? >
Yep I did test it. On stable there is nothing broken and right now even mplayer1 compiles fine. On testing there should be nothing broken apart from xbmc, where Alexis is one of upstream devs and he seems not to give fuck about making it work under both. Also Samuli broke it yesterday because he seems not to be bothered about fixing reverse dependencies with cdio update (but again it seems simple to test and fix which will be done when I am not working and have time to test it properly). So yes it works and should not pose any issues to user. I also announced it over blog to get people report more issues they find out so I can be really sure it works out. It turned out the mplayer1 really needs to work under both, which I patched yesterday for stable libav and Luca today for masked libav to work. So overall we are in green numbers if few people didn't break it on purpose or just for the ignorance. The only weird stuff might be for migrating users that they have to use "emerge -C ffmpeg && emerge -1v libav libpostproc" as a postproc is not yet split out of ffmpeg. But even that could be discussed and we can switch to split libpostproc under both libs to have matching deps (even ffmpeg has --disable-postrpoc switch :)). Tom
